Anthropic will restore access to Claude Fable 5 and Mythos 5 from July 1 after the US Department of Commerce withdrew export controls imposed over national security concerns.
Anthropic has confirmed it will begin restoring access to two of its most advanced AI models — Claude Fable 5 and Mythos 5 — starting July 1, after the US Department of Commerce officially withdrew export restrictions that had kept both models offline for nearly three weeks. The move marks the end of an unusual episode in US AI regulation that drew attention to growing government oversight of frontier AI capabilities.
What Happened: A Three-Week Suspension
The export control order was announced on June 12, after which Anthropic disabled user access to both Mythos 5 and Fable 5. The US government had flagged concerns that advanced AI models of this nature could potentially be misused by military and intelligence organisations in countries including China and Russia.
The suspension was part of a broader pattern of increased federal oversight over new AI model releases from frontier labs. As AI capabilities grow rapidly, US regulators have been working to establish mechanisms that can detect and address potential national security threats before advanced models are made widely available internationally.
On June 26, Reuters reported that the government was prepared to allow Anthropic to redeploy Mythos 5 to certain trusted US organisations beginning June 28, signalling that a broader restoration was imminent.
Anthropic’s Announcement and the Path to Restoration
Anthropic used its official account on X to confirm the development, writing that it had received notice of the lifted restrictions and would begin restoring access from Wednesday, July 1. The company said a further update would follow and thanked its users for their patience throughout the suspension period.
The restoration applies to both models, following a letter from US Commerce Secretary Howard Lutnick confirming that export controls have been formally withdrawn and that a licence is no longer required to export the Mythos and Fable models.
In the letter, accessed by Reuters, Lutnick outlined the commitments Anthropic made as part of the agreement:
- Proactively detect and address security risks associated with Fable 5 and Mythos 5
- Work with the US government on protocols, standards, and release processes for Mythos, Fable, and future models
- Inform the US government of any malicious activity detected in connection with these models
Why the US Is Tightening AI Model Oversight
The episode reflects a significant shift in how the US federal government is approaching advanced AI. As models become more capable, regulators are increasingly treating frontier AI in a similar way to other dual-use technologies — tools that have legitimate civilian applications but could also be adapted for military or intelligence purposes by adversarial states.
Anthropic, led by CEO Dario Amodei, is one of the most closely watched AI labs in this context given the capabilities of its Claude model family. The Department of Commerce’s willingness to temporarily suspend access to specific models — and then lift those restrictions after obtaining commitments from the company — suggests a developing framework for how the US government intends to manage the release of powerful AI going forward.
What This Means for Anthropic Users and Developers
For businesses, developers, and researchers who rely on Anthropic’s most capable models, the restoration of access brings an end to a period of disruption. Both Claude Fable 5 and Mythos 5 are advanced frontier models, and their absence will have affected workflows across a range of sectors.
The reinstatement also comes with a new layer of transparency around Anthropic’s obligations to US regulators. The company’s agreement to report malicious activity and collaborate on release protocols signals that future model launches from Anthropic may involve a closer coordination process with federal agencies before public deployment.
Timeline of Key Events
| Date | Event |
|---|---|
| June 12, 2025 | US Department of Commerce imposes export controls; Anthropic suspends access to Fable 5 and Mythos 5 |
| June 26, 2025 | Reuters reports US government to allow Mythos 5 redeploy to trusted US organisations from June 28 |
| June 28, 2025 | Limited Mythos 5 access restored to select trusted US organisations |
| July 1, 2025 | Anthropic begins full restoration of Fable 5 and Mythos 5 access following export control withdrawal |
What Comes Next for Anthropic and AI Regulation
Anthropic’s agreement with the Department of Commerce may set an early precedent for how frontier AI labs operate under increased federal scrutiny. The company’s commitment to proactive risk detection and government coordination on future releases — including models beyond Fable and Mythos — suggests the relationship between AI developers and US regulators is becoming more structured and formalised.
For the broader AI industry, the episode is a clear signal that the era of unrestricted model releases may be narrowing, particularly for capabilities that regulators consider to carry national security implications. How other frontier labs — and their governments — respond to similar frameworks will be worth watching closely in the months ahead.
Frequently Asked Questions
The US Department of Commerce imposed controls on Claude Fable 5 and Mythos 5 in June over concerns that advanced AI models could be misused by military or intelligence organisations in adversarial countries. It was part of the US government's broader push to tighten oversight of frontier AI releases.
Anthropic agreed to proactively identify and address security risks in its models, work with the US government on protocols for releasing Mythos, Fable, and future models, and notify federal agencies of any detected malicious activity involving these systems.
Anthropic confirmed it will begin restoring access to both models from July 1, following the formal lifting of export controls by the US Department of Commerce. A further update from Anthropic is expected once the restoration is complete.
